Transfer of regulatory T cells into abortion-prone mice promotes the expansion of uterine mast cells and normalizes early pregnancy angiogenesis

Abstract

Implantation of the fertilized egg depends on the coordinated interplay of cells and molecules that prepare the uterus for this important event. In particular, regulatory T cells (Tregs) are key regulators as their ablation hinders implantation by rendering the uterus hostile for the embryo. In addition, the adoptive transfer of Tregs can avoid early abortion in mouse models. However, it is still not defined which mechanisms underlie Treg function during this early period. Cells of the innate immune system have been reported to support implantation, in part by promoting angiogenesis. In particular, uterine mast cells (uMCs) emerge as novel players at the fetal-maternal interface. Here, we studied whether the positive action of Tregs is based on the expansion of uMCs and the promotion of angiogenesis. We observed that abortion-prone mice have insufficient numbers of uMCs that could be corrected by the adoptive transfer of Tregs. This in turn positively influenced the remodeling of spiral arteries and placenta development as well as the levels of soluble fms-like tyrosine kinase 1 (sFlt-1). Our data suggest an interplay between Tregs and uMCs that is relevant for the changes required at the feto-maternal interface for the normal development of pregnancy.

Figure 1. Number of uMCs peaks at sexual receptivity and uMCs expand with the onset of pregnancy.

Abortion-prone mice have diminished uMC numbers. (a) Number of uMCs per 1 mm² was quantified in Tolouidine blue stained paraffin sections (5 μm) from uteri of virgin CBA/J females (n = 6–8) during different phases of the estrous cycle. The mRNA levels of MC-related proteases Mcpt-1 (b), Mcpt-6 (c) and Cpa3 (d) were measured in uterine tissue of virgin CBA females at proestrus, estrus, metestrus or diestrus. (e) uMCs number per 1 mm² was quantified in uterine paraffin sections (5 μm) of BALB/c-paired CBA/J females (normal pregnant, NP) or DBA/2-mated CBA/J females (abortion-prone, AP) at gestational days (gd) 0, 2 and 5 of pregnancy. (f) During pregnancy MCs are localized in the decidua between the implantation sites. MCs (magnification ×100) were visualized after Toluidine blue O staining (0.1%) at days 2 and 5 of pregnancy in NP and AP mice and are indicated by circles. For (a,e) the results are presented as single values with means. Statistical differences were obtained using unpaired t- test (*p < 0.054, **p < 0.005, ***p < 0.001). For (b–d) the results are presented as individual values plus medians. Statistical differences were obtained using non-parametric Mann-Whitney-U- test (*p < 0.054, **p < 0.005).

Figure 2. Transfer of Tregs into abortion-prone mice is related with increase of MC numbers and reduction of fetal death rate.

(a) Fetal death rates were determined at day 12 of pregnancy in normal pregnant (NP, n = 9), abortion-prone (AP, n = 9) and Treg-treated abortion-prone (AP+Treg, n = 9) animals. (b) Representative pictures of uteri of NP, AP and AP + Treg are shown. Arrows indicates abortions. The percentage of CD4⁺Foxp3⁺ positive cells in (c) decidua (d) thymus and (e) spleen as well as the percentage of CD117⁺FcεRα⁺ positive cells in (f) decidua, (g) oviduct and (h) spleen was analyzed by flow cytometry at day 12 of pregnancy. (i) Mcpt-1 mRNA levels were analyzed by real time PCR and the data expressed as ^2−ΔcT. Data are shown as single values with medians. Statistically significant values are indicated as follows: *p < 0.054, **p < 0.005, ***p < 0.001 (Mann Whitney-U test).

Figure 3. Transfer of Tregs results in elevated numbers of MCs in abortion-prone mice and rescues placental and spiral artery defects.

Normal pregnant (NP, n = 5), abortion-prone (AP, n = 6) and abortion-prone animals transferred with Tregs (AP + Treg, n = 7) were sacrificed at day 12 of pregnancy. 5 μm cross sections of paraffin embeeded whole implantation sites (WIS) were stained with H/E. (a) Placental surface areas were measured in WIS from NP, AP and AP + Treg mice. The results are presented as single values with medians. 1 placenta was measured for each female and the median was calculated for the group. Statistical differences were obtained using non-parametric Mann-Whitney-U- test (*p < 0.054). (b) Representative pictures (magnification × 10) of Hematoxylin/Eosin stained sections from NP, AP and AP + Treg mice are shown. Broken lines indicate the placenta areas. Spiral artery (SA) wall and lumen diameters of 2 to 10 SAs per animal of NP (n = 5), AP (n = 6) and AP + Treg (n = 7) were measured and calculated as (c) wall thickness and (d) wall/lumen ratio. Data are expressed as mean ± SEM and their statistical significance was analyzed by unpaired t-test (*p < 0.054). Representative images of H/E-stained sections (magnification ×400) and immunofluorescence for smooth muscle actin and smooth muscle actin + DAPI (magnification ×1000) in SA of NP, AP and AP + Treg mice are depicted in (e) (scale bar = 20 μm).

Figure 4. Expression of placental sFlt-1 and AT₂ in abortion prone mice is normalized after transfer of Tregs.

The protein levels of the two pregnancy-relevant isoforms of soluble Flt-1 (a) (60 kDa) and (b) (145 kDa) as well as (d) angiotensin receptor II (AT2) were analyzed in placental tissue of normal pregnant (NP, n = 7), abortion-prone (AP, n = 9) and abortion-prone animals treated with Tregs (AP+Treg, n = 9) by Western blot and expressed as arbitrary units. The data are expressed as medians and analysed by Mann- Whitney-U test: p < 0.054, **p < 0.005, ***p < 0.001(Mann-Whitney-U test). (c) Representative blots for soluble Flt-1 and (e) AT2 are shown together with the housekeeping genes beta actin (for sFlt-1) and GAPDH (for AT2).